111 CHAPTER 3 SDHAF1, encoding a LYR complex-II specific assembly factor, is mutated in SDH- defective infantile leukoencephalopathy Daniele Ghezzi 1 , Paola Goffrini 2 , Graziella Uziel 3 , Rita Horvath 4,5 , Thomas Klopstock 5 , Hanns Lochmüller 5,6 , Pio D'Adamo 7 , Paolo Gasparini 7 , Tim M Strom 8 , Holger Prokisch 8 , Federica Invernizzi 1 , Ileana Ferrero 2 & Massimo Zeviani 1 1. Unit of Molecular Neurogenetics–Pierfranco and Luisa Mariani Center for Study of Children's Mitochondrial Disorders, Foundation IRCCS Neurological Institute "C. Besta", Milan, Italy. 2. Department of Genetics, Anthropology, Evolution, University of Parma, Parma, Italy. 3. Department of Child Neurology, Foundation IRCCS Neurological Institute "C. Besta", Milan, Italy. 4. Mitochondrial Research Group, Newcastle University, Newcastle upon Tyne, UK. 5. Department of Neurology, Friedrich-Baur Institute, Ludwig-Maximilians University, Munich, Germany. 6. Institute of Human Genetics, Newcastle University, Newcastle upon Tyne, UK. 7. Department of Medical Genetics, IRCCS Burlo Garofolo, University of Trieste, Trieste, Italy. 8. Institute of Human Genetics, Helmholtz Zentrum, Munich, Germany. Nature Genetics 41, 654 - 656 (2009) Published online: 24 May 2009
30
Embed
SDHAF1, encoding a LYR complex-II specific assembly factor ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
111
CHAPTER 3
SDHAF1, encoding a LYR complex-II specific assembly factor, is mutated in SDH-defective infantile leukoencephalopathy Daniele Ghezzi1, Paola Goffrini2, Graziella Uziel3, Rita Horvath4,5, Thomas Klopstock5, Hanns Lochmüller5,6, Pio D'Adamo7, Paolo Gasparini7, Tim M Strom8, Holger Prokisch8, Federica Invernizzi1, Ileana Ferrero2 & Massimo Zeviani1
1. Unit of Molecular Neurogenetics–Pierfranco and Luisa Mariani Center for Study of Children's Mitochondrial Disorders, Foundation IRCCS Neurological Institute "C. Besta", Milan, Italy. 2. Department of Genetics, Anthropology, Evolution, University of Parma, Parma, Italy. 3. Department of Child Neurology, Foundation IRCCS Neurological Institute "C. Besta", Milan, Italy. 4. Mitochondrial Research Group, Newcastle University, Newcastle upon Tyne, UK. 5. Department of Neurology, Friedrich-Baur Institute, Ludwig-Maximilians University, Munich, Germany. 6. Institute of Human Genetics, Newcastle University, Newcastle upon Tyne, UK. 7. Department of Medical Genetics, IRCCS Burlo Garofolo, University of Trieste, Trieste, Italy. 8. Institute of Human Genetics, Helmholtz Zentrum, Munich, Germany.
Nature Genetics 41, 654 - 656 (2009) Published online: 24 May 2009
112
113
We report mutations in SDHAF1, encoding a new LYR-motif
protein, in infantile leukoencephalopathy with defective succinate
dehydrogenase (SDH, complex II). Disruption of the yeast
homolog or expression of variants corresponding to human
mutants caused SDH deficiency and failure of OXPHOS-
dependent growth, whereas SDH activity and amount were
restored in mutant fibroblasts proportionally with re-expression
of the wild-type gene. SDHAF1 is the first bona fide SDH
assembly factor reported in any organism.
Succinate dehydrogenase (SDH, or complex II, cII) is composed of
four subunits (SDHA-D)1, all encoded by nuclear genes. The two
larger subunits, SDHA and SDHB, are catalytic. Dehydration of
succinate to fumarate is accomplished by SDHA through reduction of
a flavin-mononucleotide (FMN) molecule bound to its protein moiety.
This reaction is measured as succinate dehydrogenase (SDH) activity.
Electrons are then passed to three Fe-S centers bound to SDHB, which
eventually transfers them to ubiquinone (coenzyme Q, coQ). The
latter reaction is measured as succinate-CoQ reductase (SCoQR)
activity. The smaller subunits, SDHC and SDHD, anchor the complex
to the inner membrane of mitochondria. Heterozygous mutations in
SDHB, SDHC and SDHD are responsible for dominantly inherited
paragangliomas and phaechromocytomas2, 3, 4. In our series of subjects
with infantile mitochondrial disease, 22/280 (8%) had a specific
biochemical defect of cII. Nevertheless, only four 'private' mutations,
all affecting SDHA, have ever been reported, in three families with
cII-associated Leigh syndrome5, 6, 7.
114
Here we studied two family sets (Supplementary Fig. 1a online), one
consisting of a large multiconsanguineous kindred of Turkish origin
with several affected children, the other composed of three affected
children and their parents, originating from a small village in an alpine
valley of Lombardy, Italy. Two of the Italian affected children were
second-degree cousins, one being born from first-degree cousin
parents. Although we failed to formally ascertain the consanguinity of
the other parents and to connect the family of the third child with the
other two, we assumed that all affected individuals had inherited by
descent the same, presumably homozygous, mutation on the basis of
virtually identical clinical presentations and common geographic
origin.
The clinical features of the Turkish and Italian subjects were very
similar (Supplementary Table 1 online) and have partly been
described elsewhere8, 9. Symptoms consisted essentially of rapidly
progressive psychomotor regression after a 6- to 11-month disease-
free interval with lack of speech development, followed by spastic
quadriparesis and partial loss of postural control with dystonia. Brain
magnetic resonance imaging showed severe leukodystrophic changes
with sparing of the peripheral U-fibers and basal ganglia. Proton
magnetic resonance spectroscopy revealed a decreased N-acetyl-
aspartate signal and abnormal peaks corresponding to accumulation of
lactate and succinate in the white matter8, 9. Lactate and pyruvate were
variably elevated in blood. The subjects underwent relative
stabilization of their clinical conditions, with survival beyond the first
decade of life in several cases, although their growth was consistently
and severely impaired. Biochemical analysis of mitochondrial
115
respiratory chain (MRC) complexes in muscle and fibroblasts showed
a 20–30% residual activity of SDH and SCoQR, whereas the other
MRC activities were normal (Supplementary Table 2 online). Protein
blot analysis on one- and two-dimensional blue-native gel
electrophoresis showed marked reduction of cII holoenzyme in muscle
(Fig. 1a) and fibroblasts (Fig. 1b).
The methodological procedures used in the experimental workout are
reported in Supplementary Methods online. Genome-wide linkage
analysis using SNP array genotyping in the Turkish family identified a
13.5-Mb homozygous region on chromosome 19q12–q13.2 between
rs9304866 and rs2317314 with a maximal lod score of 5.7.
Concordant results were independently obtained by SNP-based
mapping of the Italian families, but here the region of continuous
homozygosity was only 1.2 Mb, between recombinant markers
rs3761097 and rs2562604, which contains 42 annotations
(Supplementary Table 3 online). A single anonymous entry in the
region, termed LOC644096, consisting of a single exon, predicts the
translation of a 115-amino-acid protein sequence (NP_001036096),
which scores high when analyzed by mitochondrial targeting
prediction programs (Supplementary Table 3). We found two
homozygous missense mutations in LOC644096—which will from
now on be termed SDHAF1, for SDH assembly factor 1—segregating
with the disease: 169G>C, corresponding to G57R in the Italian
individuals, and 164G>C, corresponding to R55P, in the Turkish
individuals (Supplementary Fig. 1b). The mutant amino acid positions
are highly conserved across species (Supplementary Fig. 1c). We
found no SDHAF1 mutations in 20 individuals with cII deficiency
116
with other clinical presentations or in 660 European and 150 Turkish
consecutive healthy control subjects.
To establish whether the SDHAF1 protein is targeted to, and resides
within, mitochondria, we expressed a hemoagglutinin-epitope (HA)-
tagged recombinant protein in COS7 cells and found that the HA-
specific immunofluorescence pattern coincides with that of mtSSB, a
mitochondrial-specific marker protein (Fig. 1c). We then found by in
vitro import assay that the SDHAF1 protein is translocated by the
proton motive–dependent transport system into the inner
mitochondrial compartment, where it is protected from digestion with
proteinase K10. The size of the in vitro translated product
corresponding to the full-length SDHAF1 gene ORF is identical to that
of the imported polypeptide (Fig. 1d), indicating that the protein does
not undergo post-import cleavage of the N-terminal mitochondrial
targeting sequence. We observed no difference in in vitro
mitochondrial translocation between wild-type and mutant SDHAF1
species (Supplementary Fig. 2 online). The SDHAF1 gene transcript is
ubiquitously expressed (Supplementary Fig. 3a online) and is
translated into a relatively hydrophilic protein with no predicted
transmembrane domain (Supplementary Fig. 3b), which suggests that
it resides in the mitochondrial matrix. This hypothesis was confirmed
experimentally by protein blot analysis on subcellular fractions of
SDHAF1HA-expressing HeLa cells. Thus, albeit essential for SDH
biogenesis, SDHAF1 is not physically associated with cII in vivo (Fig.
1e).
117
Figure 1: Protein blot and immunofluorescence studies. (a,b) One- and two-dimensional blue-native gel electrophoresis (1D- and 2D-BNGE) protein blot analysis from subject 5 (S) and a control (Ct) in muscle (a) and fibroblast homogenates (b). SDHB and COX1 are subunits of SDH (blue circle) and COX (red square) holoenzymes. (c) Confocal immunofluorescence of COS7 cells transfected with SDHAF1HA/pCDNA3.2. Scale bar, 30 m. (d) In vitro import assay. (e) Protein blot analysis of HeLa cell fractions expressing SDHAF1HA. ivT, in vitro translated SDHAF1HA; L, cell lysate; PMF, postmitochondrial fraction; Mit, mitochondrial fraction; Mem, mitochondrial membrane fraction; Mat, mitochondrial matrix. Antibodies to a mitochondrial matrix protein (ETHE1) and an inner-membrane protein complex (SDHB) were used as markers.
To test whether the disease-segregating missense mutations of
SDHAF1 are indeed causing cII deficiency, we first used a
Saccharomyces cerevisiae system. We disrupted the YDR379C-A
gene, the yeast ortholog of SDHAF1, by homologous recombination
(Supplementary Fig. 4a online). The ydr379c-a yeast strain was
OXPHOS incompetent because of a profound and specific reduction
of cII activity, whereas complex IV (cIV, cytochrome c oxidase,
118
COX) activity was normal (Fig. 2a). Transformation with the wild-
type YDR379C-A, but not with YDR379C-A variants corresponding to
the human mutant species, restored OXPHOS growth of the ydr379c-
a strain (Fig. 2a). Expression of wild-type human SDHAF1 also failed
to complement the yeast strain (Supplementary Fig. 4b), possibly
because of the low similarity between yeast and human protein
species. Respiration in standard YBN medium containing 0.6%
glucose was only slightly reduced, and cytochrome spectra were
normal (Fig. 2a), indicating the integrity of the other components of
MRC. The apparent Km value for succinate was 0.87 mM in wild-type
and 0.85 mM in the null mutant, suggesting that defective SDH
activity is caused by reduced number of enzyme units rather than by
qualitative alterations of cII. We then expressed wild-type human
SDHAF1 in three G57R mutant fibroblast cell lines. SDH and SCoQR
cII activities were completely recovered in cell line 1, whereas cell
line 2 showed partial recovery (80%) as did cell line 3 (40%) (Fig.
2b). The content of the recombinant SDHAF1 wild-type RNA was
proportional to the recovery of enzymatic activity (Fig. 2b), which
was paralleled by increased content of fully assembled cII (Fig. 2c).
Taken together, our results demonstrate that (i) mutations in SDHAF1
cause an isolated cII defect associated with a specific
leukoencephalopathic syndrome and (ii) the SDHAF1 product is the
first bona fide assembly factor specific to cII, as its loss determines
severe reduction in the amount of the enzyme in both yeast and
humans.
119
Figure 2: Complementation assays in yeast and human cells. (a) Results on YDR379C-A deleted yeast strain ( ydr379c-a). Left, OXPHOS growth. The ydr379c-a strain was transformed with wild-type YDR379C-A allele; pFL38 empty vector; and ydr379c-aR61P or ydr379c-aG63R mutant alleles. Middle, biochemical assays. Biochemical activities (nmols per min per mg protein) of SDH and COX normalized to that of citrate synthase (CS). Respiration (R) was measured as nmol O2 per min per mg dry weight. All values are expressed as percentage of the activities obtained in the control strain ydr379c-a/YDR379C-A. Right, reduced versus oxidized cytochrome spectra. Peaks at 550, 560 and 602 nm correspond to cytochromes c, b and aa3, respectively. The height of each peak relative to the baseline is an index of cytochrome content. (b) Biochemical and molecular characterization of fibroblasts cell lines. Left, SDH/CS (normal range 6.5–14.3). Middle, SCoQR/CS (normal range 8.6–18.4). Right, ratio of total/endogenous SDHAF1 mRNA. WT, wild-type control cell lines; cell lines 1, 2 and 3, G57R mutant cell lines transfected with SDHAF1/pCDNA3.2 vector; mock, G57R mutant cell line transfected with empty vector. (c) Protein blot analysis on one-dimensional blue-native gel electrophoresis. SDHB and COX1 are subunits of SDH and COX holoenzymes. N, naïve; T, transfected. SDHAF1 contains a LYR tripeptide motif, which is present in the N-
terminal region of several protein sequences in different species.
There are at least eight LYR-motif (LYRM) proteins in humans,
including SDHAF1. LYRM-4 is the human ortholog of yeast ISD11, a
protein that has an essential role in the mitochondrial biosynthesis of
120
Fe-S centers11. LYRM-6 is the 14 kDa NDUFA6 subunit of complex I
(cI)12; a second cI subunit, the 22 kDa NDUFB9 is also a LYR, iron-
responsive protein13. These data suggest that the LYR motif is a
signature for proteins involved in Fe-S metabolism. In particular,
NDUFA6, NDUFB9 and possibly SDHAF1 as well could be
important for the insertion or retention of the Fe-S centers within the
protein backbones of cI and cII, respectively. Failure of the Fe-S
centers to be incorporated within cII may eventually inhibit the
formation or destabilize the structure of the holocomplex. Although
there are other examples of low cII content and activity associated
with mutations in mitochondrial chaperonins such as yeast Tcm62.p14,
or proteins involved in Fe-S biosynthesis such as human and yeast
frataxin or IscU15, SDHAF1 is the only protein so far identified with a
specific role for cII, as other Fe-S–dependent activities were normal in
SDHAF1-defective organisms, including cI in humans and complex
III in both humans and yeast.
Acknowledgments
This work was supported by the Pierfranco and Luisa Mariani
Foundation Italy, Fondazione Telethon-Italy grant number
GGP07019, the Italian Ministry of University and Research (FIRB
2003-project RBLA038RMA), The Impulse and Networking Fund of
the Helmholtz Alliance for Mental Health in an Ageing Society, HA-
215, Deutsche Forschungsgemeinschaft HO 2505/2–1 and MIUR
grant 2006069034_003. T.K. and H.P. are members of the German
network for mitochondrial disorders (mitoNET, 01GM0862), funded
by the German ministry of education and research (BMBF, Bonn,
Germany).
121
Author Contributions
D.G. found SDHAF1 and characterized the mutations in human cells;
P.G. and I.F. carried out the experiments in yeast; G.U., R.H., T.K.
and H.L. identified the subjects and carried out the clinical workout;
P.D., P.G., T.M.S. and H.P. performed linkage analysis on the Italian
and Turkish family sets; F.I. carried out the biochemical assays on
subjects and the mutational screening on family members, disease and
healthy controls; and M.Z. conceived the experimental planning and
wrote the manuscript.
References
1. Sun, F. et al. Cell 121, 1043–1057 (2005).
2. Astuti, D. et al. Am. J. Med. Genet. 69, 49–54 (2001).
3. Niemann, S. & Muller, U. Nat. Genet. 26, 268–270 (2000).
4. Baysal, B.E. et al. Science 287, 848–851 (2000).
5. Bourgeron, T. et al. Nat. Genet. 11, 144–149 (1995).
6. Parfait, B. et al. Hum. Genet. 106, 236–243 (2000).
7. Van Coster, R. et al. Am. J. Med. Genet. 120, 13–18 (2003).
8. Bugiani, M. et al. Brain Dev. 28, 576–581 (2006).
9. Brockmann, K. et al. Ann. Neurol. 52, 38–46 (2002).
10. Neupert, W. Annu. Rev. Biochem. 66, 863–917 (1997).
11. Wiedemann, N. et al. EMBO J. 25, 184–195 (2006).
12. Cardol, P. et al. Biochim. Biophys. Acta 1658, 212–224 (2004).
13. Ye, Z. & Connor, J.R. Biochem. Biophys. Res. Commun. 275,
223–227 (2000).
14. Klanner, C., Neupert, W. & Langer, T. FEBS Lett. 470, 365–369
(a) Pedigrees of the Turkish and Italian family sets. Red numbers indicate the affected individuals who were studied. Allele genotyping symbols are +/+, for homozygous wild-type individuals, +/- for heterozygous individuals, -/- for homozygous mutant individuals. Black symbols indicate affected subjects. The dotted line indicates that the pedigree relation is suspected but not established. (b) Electropherograms of the genomic region encompassing the c.164G>C and the c.169G>C mutations (red arrows). (c) Multiple allignement of human SDHAF1 aminoacid sequence with orthologs form monkey (Macaca mulata), mouse (Mus musculus), fish (Tetraodon nigroviridis), fungi (Neurospora crassa), plant (Oryza sativa) and yeast (Saccharomyces cerevisiae), obtained with ClustalW
133 133
Figure S2
In-vitro import assay on isolated HeLa cells mitochondria of the radiolabeled full-length wild-type (wt) and mutant (containing c.169G>C or c.164G>C mutations) ORF LOC644096 products. PK: proteinase K.
134 134
Figure S3
(a) Northern-blot analysis on multiple human tissues of the SDHAF1 gene transcript encoding SDHAF1. B: brain; Pl: placenta; SM: skeletal muscle; H: heart; K: kidney; Li: liver; Lu: lung; Sp: spleen; C: colon. -actin was used as a loading control. (b) Hydropathy plot of the SDHAF1 protein sequence, generated by the TM-PRED software package. Only scores above 500 are considered as compatible with the presence of a transmembrane domain.
135 135
Figure S4.
(a) Alignment of human SDHAF1 aminoacid sequence with that of the yeast ortholog YDR379C-A. Arrows indicate the sites of mutations found in our patients and the corresponding yeast aminoacid (human R55P-yeast R61P; human G57R-yeast G63R). Accession numbers: NP_001036096 for human protein, NP_076888 for yeast protein. (b) OXPHOS growth of yeast strains in YBN medium containing 2% glucose or 2% acetate. Δydr379c-a/SDHAF1 is the Δydr379c-a transformed with human wild-type SDHAF1; Δydr379c-a/pYEX-BX is the Δydr379c-a transformed with plasmid without insert.
136 136
Supplementary table 1. Clinical features Patient / Sex Actual
age Onset First symptom Course MRI / H+MRI
1a / F 12 yrs 10 mo Acute psychomotor regression triggered by febrile illness
Severe spastic quadriplegia Growth < 3° centile Moderate cognitive impairment
Leukoencephalopathy sparing U fibers. No basal ganglia involvement Succinate peak
2a / F 10 yrs 9 mo Acute psychomotor regression
Severe spastic quadriplegia Growth < 3° centile Moderate cognitive impairment
Not performed
3 / F Died 18 mo
10 mo Acute psychomotor regression
Severe spastic quadriplegia Severe irritability Growth delay
Leukoencephalopathy sparing U fibers Succinate peak
4 / F 8 yrs 10 mo Acute psychomotor regression
Severe spastic quadriplegia Severe irritability Growth < 3° centile
Leukoencephalopathy sparing U fibers Succinate peak
5 / M 6 yrs 6 mo Deafness vomiting and slow psychomotor regression
Severe spastic quadriplegia Growth < 3° centile Stabilization and improvement after B2 therapy
Leukoencephalopathy with vacuolating process Lactate and succinate peaks
137 137
6b / F 9 yrs 10 mo Acute psychomotor regression triggered by fever
Severe spastic quadriplegia Growth < 3° centile Stabilization and improvement after B2 therapy
Leukoencephalopathy sparing U fibers. Involvement of nuclei dentati Succinate peak
7 / M 2 yrs 11 mo Acute psychomotor regression triggered by fever
Severe spastic quadriplegia Severe irritability Normal growth Stabilization and improvement after B2 therapy
Leukoencephalopathy sparing U fibers Succinate peak
a Patients previously described in Brockmann, K. et al. Ann. Neurol. 52, 38-46 (2002). b Patient previously described in Bugiani, M. et al. Brain Dev. 28, 576–581 (2006)
138 138
Supplementary table 2. Biochemical analysis of OXPHOS activities
All enzymatic activities are normalized for citrate synthase (CS) activity and expressed as percentage of the lowest control value. cII+III: succinate-cyt c oxidoreductase; nd: not determined
139
Supplementary table 3. Genes in the region obtained by homozygosity mapping
Gene Protein TargetP Mitoprot Psort Predotar PRODH2 * PROLINE
DEHYDROGENASE 2
79 78 70 16
NPHS1 KIRREL2 * kin of IRRE like 2 65 72 nd 13 APLP1 TA-NFKBH HCST TYROBP LOC728326 LRFN3 LOC644096 * hypothetical protein
ZNF260 ZNF529 ZNF382 NZNF461 LOC100129365 ZNF567 LOC342892 LOC728485 ZNF790 ZNF345 DKFZp779O175 ZNF568 * Genes that were sequenced; for these genes, the predicted mitochondrial localization (in %) of the corresponding proteins is also shown